Clinical Updates in Glycemic Control and Cardiometabolic Risk Reduction

Introduction Setting the Stage: The Controversy Robert H. Eckel, MD Effect of Earlier Glucose Control: Defining the Link Between Obesity, Diabetes, and Cardiometabolic Risk Richard W. Nesto, MD Glucose-Lowering Therapies and Their Impact on CVD Risk Silvio E. Inzucchi, MD Clinical Pearls for Achieving CVD Risk Reduction Alan S. Brown, MD, FACC

Introduction

The interaction between glycemic control and cardiometabolic risk reduction is the subject of many ongoing preclinical and clinical investigations. The data being acquired provide occasional surprises and controversy as we endeavor to unravel the multifaceted pathobiology of cardiovascular disease in our patients with diabetes and determine the ideal approach for treatment. Practitioners are challenged to keep abreast of the latest research results and assure they have access to the most relevant information as complex interactions are discovered and prevention and treatment strategies tested.

To serve this interest, Vindico Medical Education organized an educational symposium in New Orleans in November 2008, in which a prominent endocrinologist, diabetologist, and cardiologist discussed the latest clinical updates in cardiometabolic risk reduction. The disparate results of 3 major clinical trials evaluating cardiovascular outcomes in diabetic patients were introduced, creating a platform from which discussions on the link between obesity, diabetes, and cardiometabolic risk, glucose-lowering therapies and their impact on those risks, and practical advice for the practitioner could be launched. Major points from the symposium are presented in this monograph, which should provide the reader with a state-of-the-art understanding of glycemic control and cardiometabolic risk reduction.

I thank the participants for their contributions to the symposium and to the development of this monograph. Readers can be prepared to achieve a more in depth understanding of the relationship between diabetes and cardiometabolic risks and of therapeutic approaches that can benefit their patients.

Robert H. Eckel, MD
Course Director

Course Chair: Robert H. Eckel, MD Professor of Medicine Division of Endocrinology, Metabolism and Diabetes Division of Cardiology Professor of Physiology and Biophysics Professor of Pharmacy Charles A. Boettcher II Chair in Atherosclerosis Program Director, Adult General Clinical Research Center University of Colorado Denver Director, Lipid Clinic, University Hospital Aurora, CO Silvio E. Inzucchi, MD Professor of Medicine Director, Yale Diabetes Center Yale University School of Medicine New Haven, CT Alan S. Brown, MD, FACC Interventional Cardiologist Midwest Heart Specialists Director Midwest Heart Disease Prevention Center Naperville, IL Richard W. Nesto, MD Associate Professor of Medicine Harvard Medical School Chairman, Department of Cardiovascular Medicine Lahey Clinic Medical Center Burlington, MA

Setting the Stage: The Controversy

Robert H. Eckel, MD

The relationship between glucose control, diabetes, and cardiovascular disease (CVD) remains unclear, despite having been the subject of numerous large randomized clinical trials. Three recent trials provide data contributing to the controversy: ACCORD (Action to Control Cardiovascular Risk in Diabetes), ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation), and UKPDS (United Kingdom Prospective Diabetes Study). Together, these 3 studies enrolled more than 25,000 patients with diabetes and evaluated cardiovascular and glycemic control outcomes.

ACCORD

ACCORD, ADVANCE, and UKPDS contributed to the controversy surrounding the relationship between glucose control, diabetes, and CVD.   —Robert H. Eckel, MD

The ACCORD trial included 77 centers in North America that enrolled 10,251 patients with type 2 diabetes.¹ Participants averaged 62 years of age at baseline, with a median diabetes duration of 10 years. Mean baseline A1C was 8.3%, and 35% of patients had a history of macrovascular disease. Patients were randomized to receive comprehensive intensive therapy or standard therapy, with target A1C levels of <6.0% or between 7.0% and 7.9%, respectively. Treatment protocols were recommended, but no specific algorithm was mandated. The primary outcome was a composite of the first occurrence of nonfatal myocardial infarction (MI) or stroke, or death from cardiovascular causes; death from any cause was a secondary outcome. Despite achieving an average A1C of 6.4% in the intensive therapy group, compared with 7.5% in the standard therapy group, the glucose-lowering trial was terminated in February 2008 after a median follow-up of 3.4 years, approximately 17 months before the protocol-defined end of the study, because of significantly higher mortality from any cause in the intensive therapy group compared with the standard therapy group (5% vs 4%, P=.04, Table 1). In addition, although nonfatal MI was significantly greater in the standard therapy group (3.6% vs 4.6%, P=0.004) and nonfatal stroke occurrence was not different between groups (average: 1.3%), deaths from cardiovascular causes were higher in the intensive therapy group (2.6% vs 1.8%, P=.02). These surprising results have generated speculation but no discrete explanation regarding the increased deaths in the intensely controlled group. Two substudies of the ACCORD trial, one examing blood pressure control and the other examining maintenence of LDL cholesterol levels, are continuing and results remain blinded. The results of these substudies may either add to our understanding or continue the controversy.

ADVANCE

The ADVANCE study enrolled 11,140 patients with type 2 diabetes at 215 centers in 20 countries.² As in the ACCORD study, patients were randomized to receive intensive or standard therapy; however, in the ADVANCE study all intensive control patients were given the sulfonylurea gliclazide (modified release) at entry, and any standard therapy patients who were taking gliclazide (modified release) at entry were required to switch to another sulfonylurea. Additional agents were added as needed,with the same glycemic control goals as those in the ACCORD study. Patients had similar baseline characteristics compared with those in the ACCORD study: a mean age of 66 years, a diabetes duration of 8 years, an A1C of 7.5%, and a history of macrovascular disease in 32% of patients.³

Definitive clinical guidelines could not be derived from these studies due to differences among the trials.   —Robert H. Eckel, MD

Two composite endpoints comprised the primary outcome for this study: the macrovascular events that served as the composite endpoint in the ACCORD study and microvascular events including new or worsening nephropathy or retinopathy, with each endpoint assessed jointly and separately. This inclusion of microvascular events complicated the interpretation of results, especially because the association between intensive glycemic control and reduced microvascular complications is universally accepted.³

After a median 5-year follow-up, mean A1C was 6.5% in the intensive control group, compared with 7.3% in standard control group patients.² Combined micro- and macrovascular events were significantly lower in the intensive control group (P=.01); however, macrovascular events considered singly were not different between groups (P=.32). In addition, in contrast to the ACCORD results, in ADVANCE, there was no difference between the 2 groups in all-cause mortality (8.9% vs. 9.6%, P=.28, Table 1).

Table 1. ACCORD and ADVANCE:Mortality and Cardiovascular Outcomes

Sources: The ACCORD Study Group. N Engl J Med. 2008;358:2545-2459.
The ADVANCE Collaborative Group. N Engl J Med. 2008;358:2560-2572.

UKPDS

The UKPDS included 5102 patients with newly diagnosed type 2 diabetes randomized to intensive control with a sulfonylurea or with insulin, or conventional diet therapy, with drugs added as warranted.^4,5 The study was launched in 1977 with interventions through 1997, after which 3277 patients were followed for 5 years in UKPDS clinics, then for 5 additional years of follow-up using questionnaires. Patients were not asked to maintain their randomly assigned trial therapies during the post-trial follow-up. Results at the end of the intervention period in 1997, after a median follow-up of 10 years, showed that intensive blood glucose control significantly decreased the risk of microvascular complications and any diabetes-related endpoint, with a relative risk reduction (RRR) of 25% (P=.0099) and 12% (P=.029), respectively (Table 2). However, the RRR for MI and all-cause mortality (16%, P=.052, and 6%, P=.44, respectively) were not statistically significant. Within 1 year following discontinuation of the intervention, the difference in glycemic control between the intensive and conventional treatment groups had disappeared; however, in 2007, after a median of 8.5 years of post-intervention monitoring, diabetes-related endpoints and microvascular disease outcomes remained significantly improved. The 15% RRR for MI and 13% RRR for all-cause mortality were statistically significant (P=.014 and P=.007, respectively).

In summary, 3 major multicenter randomized trials provided disparate results with regard to the relationship between glycemic control and occurrence of macrovascular events. Many similarities among the studies existed, but differences, particularly in the length of follow-up (median follow-up 3.4, 5.0, and 10.0 years), preclude making conclusions or deriving definitive clinical guidance from these studies alone.

Table 2. Effect of Glucose Control Maintained through 1997, with Post-Trial Follow-up

Source: Holman RR, et al. N Engl J Med. 2008;359:1577-1589.

References

1. The ACCORD Study Group. N Engl J Med. 2008;358:2545-2459.
2. The ADVANCE Collaborative Group. N Engl J Med. 2008;358:2560-2572.
3. Dluhy RG, McMahon GT. N Engl J Med. 2008;2630-2633.
4. UKPDS Group. Lancet. 1998;352:837-853.
5. Holman RR, et al. N Engl J Med. 2008;359:1577-1589.

Effect of Earlier Glucose Control: Defining the Link Between Obesity, Diabetes, and Cardiometabolic Risk

Richard W. Nesto, MD

Obesity, diabetes, metabolic syndrome, and insulin resistance provide a continuum of risks for cardiometabolic disease. Structural and functional abnormalities of the cardiovascular system have been reported in all these conditions. The complexity of the mechanisms responsible for these changes (and particularly the effect of glycemia) may be exemplified in phenomena such as the “legacy effect” that was reported in the UKPDS (United Kingdom Prospective Diabetes Study). In that study, despite a convergence of glycemic control after interventions were discontinued, an emergent cardioprotective effect was observed.¹ Data from many recent studies not only have increased our understanding of these interrelationships, but also have fueled the controversy associated with cardiovascular risk factors and their management.

Treating to New Targets (TNT) – Lowering LDL and Metabolic Syndrome

TNT was a major trial of 10,001 patients with clinically evident coronary heart disease (CHD) to investigate the effects of lowering low-density lipoprotein (LDL) cholesterol to <100 mg/dL.^2,3 Patients were randomized to receive atorvastatin 10 mg/day or 80 mg/day and were followed for a median of 4.9 years. The high-dose group had a 22% reduction (P<.001) in major cardiovascular events compared with those on low-dose atorvastatin. Patients in the high-dose subgroup with metabolic syndrome and diabetes also had reduced risk compared with similar patients on the low dose; however, there was still substantial residual risk in these patients despite maximum LDL-lowering to an average of 77 mg/dL. In addition, patients with metabolic syndrome and diabetes on 80 mg of atorvastatin had greater residual risk for major cardiovascular events than similar patients without diabetes.

Vascular homeostasis requires a balance of several biochemical factors and is disrupted by obesity, type 2 diabetes, and insulin resistance (Figure 1). Five metabolic factors associated with residual risk include some of the criteria used to diagnose metabolic syndrome, and these factors are not unrelated: increased waist circumference (men ≥40 inches; women ≥35 inches), elevated fasting glucose (≥100 mg/dL), reduced high density lipoprotein (HDL) (men <40 mg/dL; women <50 mg/dL), elevated triglycerides (≥150 mg/dL), and elevated blood pressure (≥130/85 mm Hg).⁴ These 5 metabolic factors represent additional targets for intervention that may reduce residual risk for MI. At present, significant attention is being directed to evaluating the effects of lowering fasting glucose to remove some of this residual risk; however, based on the studies reviewed earlier in this monograph, there is indecision regarding its potential value.

Figure 1. Beyond High Cholesterol and Hypertension: Factors Responsible for Atherosclerosis and the Induction of Acute MI in Diabetes Vascular homeostais requires a balance of several biochemical factors.

Obesity and Cardiovascular Risk

Fat varies both histologically and physiologically, depending on its location in the body and association with other organ systems. For example, visceral and hepatic fat are similar to the fat around the heart and that which lines the vasculature. It is also histologically similar to fat around the waist, and there is a linear relationship between waist circumference and the amount of fat around the heart; people with larger waists have more adipose tissue lining their blood vessels.⁵

This process starts in childhood. In a study of 50 obese children aged 12 years, they had structural changes in their carotid vessels including increased intima-medial thickness (IMT) compared with age and sex-matched controls.⁶ In this study, there was also a direct positive relationship between carotid IMT and other indices of insulin resistance, including fasting glucose to insulin ratio (FGIR), homeostasis model assessment of insulin resistance (HOMA-IR), and quantitative insulin-sensitivity check index (QUICKI).

Other evidence related to the early emergence of risk factors in association with obesity was provided in a study of 45 obese children with a mean age of 15 years.⁷ Obese children had significantly elevated levels of matrix metalloproteinase-9 (MMP-9), a marker that correlates with both future MI risk and unfavorable outcome in the event of an MI, and tissue inhibitor-metalloproteinase-1 (TIMP-1). Values of MMP-9 and TIMP-1 correlated with blood pressure, muscle mass index, and fasting insulin levels.

Finally, a study of the arteries of more than 1000 men aged 15 to 24 years who died of nonmedical causes were shown to have weight-dependent fatty streaks and raised lesions in the aorta and right coronary artery.⁸ These results, although not seen in similar-aged women, indicate that obesity in adolescents and young adults accelerates the progression of atherosclerosis before clinical manifestations appear.

Coronary Vasculature - Pathobiology Studies

Basic research in animal models has shown that, after being fed a high-fat diet for 7 weeks, periadventitial adipose tissue increased in proportion to increases in the animal’s weight. Furthermore, it was shown in obese humans that the interface between the perivascular adipose tissue and adventitia of an atherosclerotic aorta is infiltrated with inflammatory cells, most likely related to the strong chemotactic activity of cytokines secreted by the adipose tissue.⁹

An intracoronary ultrasound study in humans revealed a significantly higher percentage of lipid area (38 ± 19% vs. 30 ± 19%, P=.02) and lipid volume (39 ± 17% vs. 33 ± 17%, P=.03) in the plaques of patients with metabolic syndrome, compared with the plaques of patients without metabolic syndrome, even though stenosis severities were the same in the 2 groups.¹⁰ In a multivariate analysis after adjusting for confounding and non-metabolic syndrome coronary risk factors, metabolic syndrome was an independent predictor of lipid-rich plaque (odds ratio 4.00, P=.01). In a similar study in patients with diabetes, impaired glucose regulation (IGR), or normal glucose regulation showed with logistic regression after controlling for confounding and coronary risk factors that diabetes and IGR were significantly (P=.03) associated with a more lipid-rich plaque.¹¹ The authors concluded that this may be related to increased insulin resistance in these patients. Other data suggest that waist circumference determines the composition of lipid fraction more than other risk factors for metabolic syndrome.

Similar to the inflammatory response in plaque of obese patients, the cellular composition of carotid plaques removed at surgery from patients with diabetes differs from that of patients without diabetes, including the presence of localized dense infiltrations of mononuclear cells that express surface markers associated with inflammation.¹² The multiligand cell surface receptors for advanced glycosylated endproducts (RAGE) are present in these plaque regions in concentrations that are related to the patient’s glycemic control. Attachment of the advanced glycosylated endproducts (AGEs) to the receptors induces a chemotactic response and enhances the inflammatory reaction.AGEs are deposited in the shoulder regions, where the plaques are known to disrupt. Staining reveals localization of the receptor, inflammatory cells, and MMP-9 in this area, which is also dependent on glycemic control. These data suggest that diabetic patients are not only more likely to have advanced atherosclerosis, but also, by being more inflammatory, have more dangerous atherosclerosis.

The effect of obesity on cardiovascular risk factors is reversible.   —Richard W. Nesto, MD

Other changes in cardiovascular function that have been reported in patients with abnormal glucose regulation include myocardial blood flow, which was measured in subjects who were insulin sensitive (IS), were insulin resistant (IR), had impaired glucose tolerance (IGT), were normotensive with type 2 diabetes (DM), or were diabetics with hypertension (HTN).¹³ Total vasodilator capacity during pharmacological vasodilation and vasomotion in response to cold pressor testing were evaluated in each group. Total vasodilator capacity was similar in the 3 normoglycemic groups (IS, IR, and IGT) and significantly decreased in DM patients (-17%) and HTN DM patients (-34%). Vasomotion, however, was significantly decreased in IR (-56%), IGT (-85%), DM (-91%), and HTN (-120%) patients compared with IS controls. These results demonstrated a loss of coronary flow and function with impaired glucose control or HTN.

In addition, myocardial triglyceride content, quantified by magnetic resonance imaging and spectroscopy studies on 134 patients, was shown to increase with progression from leanness to obesity to IGT to diabetes (Figure 2).¹⁴ Diastolic function, measured by peak filling rate, was decreased significantly in patients who were obese (79 mL/ms), with IGT (63 mL/ms), and with diabetes (71 mL/ms), compared with lean (108 mL/ms) patients. These results are consistent with the observation that diastolic heart failure is more prevalent in diabetics than in non-diabetics and that cardiac lipotoxicity is a factor in cardiomyopathy in obesity-related disorders.

Cardiovascular Risk Factor Intervention Trials

The effect of obesity on diastolic function is reversible, as shown by a study in which 12 obese diabetic patients lost an average of 8 kg on a 16-week very low calorie diet.¹⁵ In addition to losing myocardial and hepatic fat, lowering triglycerides, and achieving a significant improvement in glycemic control (A1C decrease from 7.9% to 6.3%, P=.006), patients experienced a significant improvement in diastolic function (increased ratio between early and atrial filling phases: 1.02 to 1.18, P=.019).

The reversibility of risk factors was further demonstrated when more rigorous measures were taken to achieve weight loss. In the prospective, non-randomized Swedish Obese Subjects Study, patients undergoing gastric surgery were matched with those undergoing conventional treatment.¹⁶ At the 10-year follow-up, rates of recovery from hypertriglyceridemia (46% vs. 24%), low HDL cholesterol (73% vs. 53%), diabetes (36% vs. 13%), hypertension (19% vs. 11%), and hyperuricemia (48% vs. 27%) were significantly greater in the surgery group compared with conventional treatment subjects (P<.05), and weight loss in the surgery group was 16.1%, compared with an increase of 1.6% in control subjects. At the 16-year follow-up, mortality was also lower in the surgical group (hazard ratio 0.76, 95% CI 0.59-0.99; P=.04).¹⁷

Figure 2. Cardiac Steatosis in Prediabetic Humans: A Spectrum of Alterations form Obesity to DM Myocardial tryglyceride content increases during the progression from leanness to obesity to impaired glucose tolerance to diabetes. Source: McGavock JM, et al. Circulation. 2007; 116:1170-1175.

Evaluating intensive treatment in patients with additional risk factors, the Steno-2 trial randomly assigned 160 patients with both diabetes and microalbuminuria, who were therefore at high risk for cardiovascular disease, to either intensive treatment for diabetes to lower cholesterol and blood pressure accompanied by extensive lifestyle advice, or conventional treatment.^18,19 An intervention period of 7.8 years was followed by a 5.5-year observational period. The proportion of patients who achieved the glycemic control target of A1C <6.5% did not reach significance between the 2 groups, and the proportions of patients who achieved target goals for triglycerides and diastolic blood pressure were also not different. Nevertheless, a significantly greater proportion of patients in the intensive therapy group had lower cholesterol and systolic blood pressure (Figure 3), and fasting glucose decreased significantly in the intensive therapy group from 178 mg/dL to 129 mg/dL and A1C decreased from 9.0% to 7.9% (P<.01). Starting at 9 years, additional significant differences in favor of the intensive therapy group emerged: deaths from all causes (P=.02), cardiovascular deaths (P=.04), and incidence of cardiovascular events (P<.001). In this study, the reduction in total cholesterol was responsible for approximately 70% of the cardiovascular benefit.²⁰

In the recent 3-year SANDS (Stop Atherosclerosis in Native Diabetics Study), fewer patients in the intensive target group (LDL=70 mg/dL and systolic blood pressure=115 mm Hg) had progression of carotid intima-medial thickness (CIMT), and there was a trend toward fewer increases in left ventricular (LV) mass compared with the standard target group (LDL=100 mg/dL and systolic blood pressure=130 mm Hg). ²¹ The results of this study suggested that lowering LDL and systolic blood pressure beyond what is currently recommended could provide additional cardiovascular benefits.

Finally, the ongoing BARI 2D (Bypass Angioplasty Revascularization Investigation 2 Diabetes) trial is evaluating patients with diabetes and stable coronary artery disease (CAD) who were randomly assigned to either an insulin-providing regimen with sulfonylurea and insulin, or an insulin-sensitizing regimen with metformin and rosiglitazone.²² Patients were randomly assigned to revascularization of choice or intensive medical treatment for angina. Results of this study are expected to further address whether the approach to treating diabetes can have an impact on cardiovascular outcomes.

In conclusion, relationships are being established that assist with understanding and addressing this complex clinical challenge. Importantly, prevention and intervention should be introduced early in the process, and research should continue to assure optimum therapeutic regimens are available for the entire spectrum of risk factors.

Figure 3. Target Risk Factor Attained with Intensive Treatment Program at 7.8 Years Patients undergoing intensive treatment experienced greater improvements in risk factors compared to those receiving conventional treatment Source: Gaede P, et al. N Engl J Med. 2003;348:383-393.

References

1. Holman RR, et al. N Engl J Med. 2008;359:1577-1589.
2. LaRosa JC, et al. N Engl J Med. 2005;352:1425-1435.
3. Deedwania P, et al. Lancet. 2006;368:919-928.
4. Grundy SM. J Clin End Metab. 2004;89:2595-2600.
5. Yudkin JS, et al. Lancet. 2005;365:1811-1820.
6. Atabek ME, et al. Pediatr Red. 2007;61:345-349.
7. Glowinska-Olszewska B, et al. Metabolism. 2007;56:799-805.
8. McGill HC Jr, et al. Circulation. 2002;105:2712-2718.
9. Henrichot E, et al. Arterioscler Thromb Vasc Biol. 2005;25:2594-2599.
10. Amano T, et al. J Am Coll Cardiol. 2007;49:1149-1156.
11. Amano T, et al. J Am Coll Cardiol: Cardiovascular Imaging. 2008;1:39-45.
12. Cipollone F, et al. Circulation. 2003;108:1070-1077.
13. Prior JO, et al. Circulation. 2005;111:2291-2298.
14. McGavock JM, et al. Circulation. 116:1170-1175.
15. Hammer S, et al. J Am Coll Cardiol. 2008; 52:1006-1012.
16. Sjostrom L, et al. N Engl J Med. 2004;351:2683-2693.
17. Sjostrom L, et al. N Engl J Med. 2007;357:741-752.
18. Gaede P, et al. N Engl J Med. 2008;358:580-591.
19. 19.Gaede P, et al. N Engl J Med. 2003;348:383-393.
20. Gaede P, Pedersen O. Diabetes. 2004;53(Suppl 3):S39-S47.
21. Howard BV, et al. JAMA. 2008;299:1678-1689.
22. Brooks MM, et al. Am J Cardiol. 2006;97:9G-19G.

Glucose-Lowering Therapies and Their Impact on CVD Risk

Silvio E. Inzucchi, MD

During the past 10 to 15 years, treatment options for patients with diabetes have increased dramatically. This can be challenging to the practitioner who must determine for the individual patient which drug is best, what combinations may be effective, and which drugs must be avoided.

Understanding treatment choices requires understanding the biologic consequences of the pathogenesis of type 2 diabetes: insulin resistance and decreased insulin supply (Figure 1). Insulin resistance is first demonstrated in the muscle, which reduces glucose uptake from the circulation. Thus, the pancreas responds to the elevated circulating glucose levels by producing and secreting more insulin, resulting in hyperinsulinemia. Normally, the liver stops producing glucose in response to insulin; however, in the setting of hepatic insulin resistance, endogenous glucose production continues, resulting in both fasting and postprandial hyperglycemia. “Pancreatic exhaustion” can occur, where the pancreas is unable to continue to produce enough insulin to compensate for the insulin resistance in peripheral tissues. Finally, the intestine is also involved in the process, both as a conduit of carbohydrate calories and also as the native site of the incretin system, further contributing to the pathogenic complexity of type 2 diabetes.

In the progression from insulin resistance to diabetes, the hyperglycemia that results as the pancreas loses its ability to provide the necessary increase in insulin is manifested first postprandially and then in the fasting state. Microvascular complications such as neuropathy, retinopathy, and nephropathy typically are not diagnosed until the patient actually has diabetes. Macrovascular complications, specifically myocardial infarction (MI) and stroke, and to some extent heart failure, can occur throughout the process from insulin resistance to impaired glucose tolerance to diabetes. These are important considerations for the practitioner when managing patients with cardiovascular risk factors.

Figure 1. Pathogenesis of T2DM and Drug Class Targets Therapeutic agents for type 2 diabetes target many specific aspects of glucose regulation.

Current Treatment Strategies for Type 2 Diabetes

Treatment strategies for type 2 diabetes include lifestyle modification and therapeutic agents that target specific aspects of abnormal glucose regulation (Figure 1; Table 1). Diet and exercise are fundamental mainstays of diabetes treatment, which contribute to decreasing insulin resistance. Pharmaceutical therapies include several classes that increase insulin, either by enhancing endogenous production or by providing exogenous insulin.

Table 1. Type 2 Diabetes Treatment Strategies: 2008

Sulfonylureas

The sulfonylureas, which increase insulin secretion by the pancreas, have been in use for more than 50 years and have moderate efficacy, decreasing A1C 1% to 2%. Although they are inexpensive, their main disadvantages include hypoglycemia and weight gain. In the United Kingdom Prospective Diabetes Study (UKPDS), although both sulfonylureas and insulin significantly reduced microvascular complications, macrovascular events were not significantly reduced compared with conventional diet therapy.¹ A “legacy effect” of intensive glucose control, however, was seen as a 15% (P=.01) reduction in MI at the 10-year follow-up of this study despite the loss of between-group differences in A1C within the first post-trial year.²

Nonsulfonylurea secretagogues

The glinide mechanism of action is similar to that of the sulfonylureas, although their rapid on/off action is more similar to the normal physiology of insulin secretion. They have moderate A1C reduction efficacy (1% to 1.5%) and target postprandial glucose. Disadvantages include weight gain, hypoglycemia, and frequent dosing.

Incretins

Incretins are hormones made by the gastrointestinal tract and include glucagon-like peptide 1 (GLP-1), which is produced by neuroendocrine cells. GLP-1 induces the pancreas to increase insulin production and secretion and modulates glucagon secretion, both in a glucose-dependent fashion. In addition, it decreases the rate of gastric emptying and enhances satiety. Exenatide, a synthetic drug based on GLP-1, has a beneficial effect on islet cell function, therefore increasing insulin levels and decreasing glucagon. Beta cell preservation has been demonstrated in animal models and is not yet confirmed in humans. Exenatide typically produces a 1% decrease in A1C. A primary benefit of exenatide is weight loss, ranging from 5 lb to 15 lb, which is considerably higher than that achieved with the biguanide metformin. Because it has little effect on insulin secretion when blood glucose is normal, hypoglycemia is rare, a potentially important safety advantage. Disadvantages include administration by injection, although promising results are being obtained with a once-weekly formulation compared with the current twice-daily regimen. Side effects include nausea and occasional vomiting. There is no definite evidence of a causal link between GLP-1 analogues and pancreatitis, which has been recently reported.

In addition to its effect on glycemic parameters, beneficial cardiovascular effects have been reported with exenatide, although mainly in open-label extension studies.³ Significant improvements from baseline were observed for A1C, total cholesterol, triglycerides, LDL cholesterol, systolic and diastolic blood pressure, and body weight (Table 2).⁴ GLP-1 receptors are demonstrated in cardiomyocytes, and preliminary data from very small studies suggest that GLP-1 may have cardioprotective effects extending to improved peripheral blood flow, left ventricular function, and exercise capacity, and decreased arrythmia.^5-8

Table 2. GLP-1 Agonist, Exenatide: Effect on Cardiovascular Risk Factors

Source: Klonoff DC, et al. Curr Med Res Opin. 2008;24:275-286.

Dipeptidyl Peptidase-4 Inhibitors

Dipeptidyl peptidase-4 (DPP-4) inhibitors are a new drug class in the diabetes armamentarium. The enzyme DPP-4 rapidly inactivates incretins, resulting in very short half-lives of 2 to 3 minutes for these peptides in vivo. DPP-4 inhibitors, therefore, increase levels of the endogenous incretins. Sitagliptin, currently the only DPP-4 inhibitor approved for the treatment of diabetes in the United States, has modest efficacy on A1C, with a 0.6% to 0.8% reduction - although, as with other agents, this is to some extent dependent on baseline A1C levels, with a larger effect demonstrated typically in those with the highest baseline glycemia. Consistent with its mechanism of action, sitagliptin is not associated with hypoglycemia, and data regarding beta cell preservation are currently available only from animal studies. It is well-tolerated, with some reports of urticaria and rare cases of angioedema that have not been established as being drug-related.

Biguanides

The biguanide metformin decreases hepatic glucose production and is the most commonly used drug for type 2 diabetes worldwide. Its efficacy is similar to that of the sulfonylureas; that is, patients experience a 1% to 2% decrease in A1C on therapy. Early in the treatment phase, patients usually lose weight or its effect is weight-neutral. Because its mechanism of action is not on pancreatic insulin secretion, it is not associated with hypoglycemia; however, approximately 50% of patients report abdominal pain, cramping, or diarrhea, but treatment needs to be discontinued in only about 5% of patients. Lactic acidosis occurred in 1 in 30,000 patients and almost exclusively in patients contraindicated because of renal failure. Metformin should also be used cautiously in the very elderly patients, due to frequently coexistant and often unrecognized renal insufficiency. In the UKPDS, the small subgroup of overweight patients receiving metformin had a significant reduction in MI (39%; P=.01) and coronary deaths (50%; P=.02) as compared to diet therapy.⁹ The “legacy effect” of intensive glucose control was also seen in metformin patients, with a 33% (P=.005) reduction in MI at the 10-year follow-up, despite the loss of between-group differences in A1C in the first post-trial year.²

Alpha Glucosidase Inhibitors

Alpha glucosidase inhibitors (AGI) are weight-neutral drugs that decrease carbohydrate absorption in the GI tract, targeting postprandial glucose. They are rarely used in the United States because of their limited efficacy (0.5% to 1.0% A1C reduction), significant GI toxicity, and frequent dosing. They may, however, have some cardiovascular benefit. In the STOP-NIDDM study of reducing cardiovascular disease risk in prediabetic patients, the AGI acarbose reduced relative risk for coronary heart disease, cardiovascular death, congestive heart failure, and peripheral vascular disease in patients with impaired glucose tolerance by 49% compared with patients taking placebo (P=.04).¹⁰

Thiazolidinediones (TZDs)

The TZDs, or glitazones, increase glucose uptake in the skeletal muscle in the presence of insulin, with a typical reduction in A1C of 1% to 1.5%. Also, they are not associated with hypoglycemia, and studies have demonstrated a beneficial effect on beta cell viability, which affords greater glycemic control durability. Many cardiovascular benefits have been associated with these drugs that include increased HDL cholesterol and decreased C-reactive protein, carotid intima-medial thickness, and coronary atherosclerosis; however, recent concern about cardiovascular risks have aroused controversy. Use of the drugs in predisposed patients, which may include diabetic patients with systolic and/or diastolic dysfunction, may precipitate clinical heart failure. Such an effect appears not to be due to any deleterious effect on ventricular function, but instead relates to sodium retention induced at a renal level. Edema is therefore another side effect of TZDs, as is weight gain, especially if used with insulin. These agents also have a slow onset of action. Emerging data suggest there may be an increased fracture risk in female patients. On the other hand, data supporting improved cardiovascular outcomes were reported from the randomized PROactive (Prospective Pioglitazone Clinical Trial in Macrovascular Events) study.^11-13 The cardiac paradox of TZD therapy became more complicated by recent controversial meta-analyses suggesting that rosiglitazone, previously considered to reduce atherosclerosis, may actually increase the risk of MI. Subsequent data from large trials, including ACCORD (Action to Control Cardiovascular Risk in Diabetes), cfailed, however, to support an association of rosiglitazone with adverse cardiovascular outcomes. Other studies addressing the issue of TZDs and cardiovascular effects are underway.

Figure 2. Algorithm for the Metabolic Management of Type 2 Diabetes The ADA and EASD released a new consensus statement in November 2008 for the treatment of patients with type 2 diabetes. Source: Nathan DM, et al. Diabetes Care. 2009; 32:193-203.

American Diabetes Association Consensus Statement: November 2008

In November 2008, the American Diabetes Association (ADA), in conjunction with the European Association for the Study of Diabetes, released a new consensus statement on the treatment of patients with type 2 diabetes.¹⁴ The recommendations are presented in an algorithm that includes metformin as the initial pharmacotherapy, followed by well-validated core therapies comprising either a sulfonylurea or basal insulin. The alternative path, a Tier 2 therapy, includes the “less well-validated therapies” pioglitazone or the incretin GLP-1 agonist exenatide.

In summary, there is strong evidence demonstrating a beneficial effect of antihyperglycemic drugs on microvascular endpoints, supporting the longstanding glucose hypothesis that hyperglycemia is either directly or indirectly related to the development or progression of diabetes-specific complications including retinopathy, nephropathy, and neuropathy. There is still much to be learned, however, about the cardiovascular effects of most antihyperglycemic agents. Results from several recent studies suggest that, if a benefit exists, it is important that the treatment program be applied early in the process; especially in those with shorter duration of diabetes and less-established coronary artery disease. In addition, it may take years for the benefits to develop. Other than for heart failure in certain at-risk patients treated with TZDs, there are currently no proven positive or negative cardiovascular effects of any specific drug or drug combination.^15-17

References

1. UKPDS Group. Lancet. 1998;352:837-853.
2. Holman RR, et al. N Engl J Med. 2008;359:1577-1589.
3. Bergenstal R, et al. AHA. 2008; Poster P27.
4. Klonoff DC, et al. Curr Med Res Opin. 2008;24:275-286.
5. Nystrom T, et al. Am J Physiol Endocrinol Metab. 2004;287:E1209-1215.
6. Nikolaidis LA, et al. Circulation. 2004;109:962-965.
7. Sokos GG, et al. J Cardiol Failure. 2006;12:694-699.
8. Sokos GG, et al. Am J Cardiol. 2007;100:824-829.
9. UKPDS Study Group. Lancet. 1988;352:854-865.
10. Chiasson JL, et al. JAMA. 2003;290:486-494.
11. Dormandy JA, et al. Lancet. 2005;366:1279-1289.
12. Erdmann E, et al. J Am Coll Cardiol. 2007;49:1772-1780.
13. Wilcox R, et al. Stroke. 2007;38:865-873.
14. Nathan DM, et al. Diabetes Care. 2009;32:193-203.
15. ACCORD Study Group. N Engl J Med. 2008;358:2545-2559.
16. ADVANCE Collaborative Group. N Engl J Med. 2008;358:2560-2572.
17. Duckworth WC. Scientific Sessions of the American Diabetes Association. San Francisco, June 2008.

Clinical Pearls for Achieving CVD Risk Reduction

Alan S. Brown, MD, FACC

90% of weight loss is calorie restriction 90% of weight maintenance is exercise...and eating breakfast every day.   —Alan S. Brown, MD, FACC

The prevalence of diabetes in the United States has increased 5-fold since the 1950s and is associated with increased calorie consumption and body weight. In fact, the increasing body mass index (BMI) of the American population correlates with the increased incidence rate of type 2 diabetes. However, the rise in obesity and diabetes is occurring worldwide as consumption of higher calorie foods becomes more prevalent. When managing patients with cardiometabolic risk factors, practitioners must consider that many drugs may aggravate obesity and, when possible, substitute weight neutral and weight reducing alternative medications for drugs that may promote weight gain.^1-4 Health care providers should work together to derive the most appropriate therapeutic regimen for the individual patient.

Weight Loss and Weight Maintenance

When working with overweight patients, the initial emphasis should be on calorie restriction. 90% of weight loss is achieved through calorie restriction. Exercise is important for weight maintenance and decreases the loss of lean body mass when combined with diet. Patients benefit from being referred to weight loss programs that have a proven track record of results. Several commercial weight loss programs have been shown to be effective. Once weight loss is achieved, there are 2 common characteristics among people who are successful at maintaining their weight: exercise, which is critical for weight maintenance, and the habit of eating breakfast every day.

In patients with metabolic syndrome, a 10% reduction in weight provides for a 56% reduction in risk for developing type 2 diabetes. This information can be useful to help patients set realistic weight loss expectations. Patients are often surprised to hear that such a modest reduction in weight is not only very achievable but provides significant health benefits.

Diabetes and Cardiovascular Disease Risk

Diabetes and cardiovascular disease are well-correlated.⁵ More than 75% of hospitalizations for diabetic complications are for cardiovascular issues, with congestive heart failure (CHF) the most common diagnosis. This is partly because insulin resistance and heart failure comprise a feed-forward cycle: insulin resistance leads to heart failure, and heart failure leads to insulin resistance.

The landmark East West Study compared 7-year myocardial infarction (MI) incidence in patients with diabetes and no history of MI to patients without diabetes and a history of MI (Figure 1).⁶ The hazard ratio (HR) for death from coronary heart disease for the 2 groups was not significantly different from 1.0 (HR 1.4; 95% CI, 0.7-2.6); ie, a person with diabetes without prior coronary disease has the same risk of dying from MI as a person with a prior coronary event but without diabetes. The 7-year incidence of MI was significantly greater in nondiabetic patients with a previous MI (18.8%) compared with those without a history of MI (3.5%). The incidence was much greater in corresponding patients with diabetes, and diabetic patients with a prior MI had more than twice the 7-year MI rate of patients without a prior MI (45.0% vs. 20.2%, P<.001). These data support that treatment of cardiovascular risk factors in patients with diabetes and no history of MI should be as aggressive as in patients without diabetes but with a history of previous MI.

Simple Steps to Reduce Cardiovascular Events in Patients with Type 2 Diabetes

Six therapeutic steps can be taken to reduce cardiovascular (CV) events in all patients with type 2 diabetes: aspirin, statin (regardless of baseline LDL cholesterol levels) angiotensin-converting enzyme (ACE) inhibitor/angiotensin receptor blocker (ARB) in all patients with at least 1 additional CV risk factor, Β-Blocker in all patients with CAD/congestive heart failure (CHF) or history of MI (carvedilol, nebivolol are preferred since they don’t aggravate insulin resistance), diet and exercise, and aggressive glucose management.

Figure 1. Type 2 Diabetes and CHD 7-Year Incidence of Fatal/NonFatal MI (East West Study) A person with diabetes without prior coronary disease has the same risk of dying from MI as a person with a prior coronary event but without diabetes. Source: Haffner SM, et al. N. Engl J Med. 1998;339:229-234.

Emerging Risk Factors

Understanding relevant cardiovascular risk factors is essential for managing the diabetic patient. A maxim holds that any effective preventive therapies should yield the most benefit in patients at the highest risk. Accordingly, if we can find novel methods of assessing risk that may indicate subclinical atherosclerosis, we may have an opportunity to intervene earlier with proven therapies and hence have a greater impact on progression of disease. Three emerging risk factors are the subject of investigation: carotid intima-medial thickness (CIMT), coronary artery calcium deposits, and high-sensitivity C-reactive protein (hs-CRP).

CIMT

The ENHANCE (Ezetimibe and Simvastatin in Hypercholesterolemia Enhances Atherosclerosis Regression) study randomly assigned 720 patients to receive simvastatin or simvastatin and ezetimibe, a drug with action complementary to the statins that provides a further boost to LDL cholesterol reduction.⁷ The 2-year study failed to show a difference between groups in IMT, the primary endpoint, importantly because baseline CIMT values were normal, which was considered to be a serious flaw in the study.⁸ LDL cholesterol reduction in the ezetimibe group was 17%; however, whether it is sufficient to improve outcomes was unclear from this study.

The primary endpoint of CASHMERE (Carotid Atorvastatin Study in Hyperlipidemic Post-Menopausal Women — a Randomized Evaluation of Atorvastatin vs. Placebo) was also change in CIMT. A total of 339 postmenopausal patients were enrolled in the 2x2 factorial design study that randomized patients to atorvastatin and hormone replacement therapy, alone or in combination, versus placebo.⁹ Baseline IMTs were normal in both the atorvastatin and placebo groups as they were in ENHANCE, and at the end of the 12-month treatment interval, there was no significant between-group difference in CIMT, despite significant differences in lipid profile. Other cardiovascular secondary endpoints were also not significantly different between atorvastatin and placebo groups.¹⁰ Many clinical trials with atorvastatin have proven that it is an effective agent in reducing CV events and hence CIMT studies, especially if performed on patients with normal baseline CIMT, may not be the ideal predictor of future CV event reduction.

Regardless of the negative results of these studies, scientific review and further appropriately designed studies are necessary to reach conclusions regarding the validity of CIMT progression as a surrogate for cardiovascular events.

Coronary Artery Calcium Deposits

High coronary artery calcium deposit scores, which can be measured by electron beam computed tomography (EBCT), have been correlated with a high incidence of cardiovascular events; in fact, some studies have suggested that EBCT may provide risk information beyond that of Framingham risk scoring for predicting development of atherosclerosis. In addition, it is often less expensive than a traditional stress test. EBCT may be a valid option for a borderline patient, for example, one whose LDL is 170 mg/dL with a goal of 160 mg/dL, but who is resistant to taking medication. If the results of an EBCT are positive, then the patient should be treated aggressively. Other candidates for EBCT include patients with a family history of disease who, in the presence of a substantial calcium score bon the EBCT, should be treated aggressively, targeting an LDL <100 mg/dL.

HS-CRP

The JUPITER (Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin) trial randomized 15,000 healthy men aged ≥50 years and women aged ≥60 years with LDL <130 mg/dL, but with a CRP level ≥2 mg/mL, to receive rosuvastatin or placebo.¹¹ The study was stopped after 2 years of the planned 4-year intervention period, when interim data analysis revealed the rosuvastatin group had a significant reduction in the composite primary endpoint of cardiovascular morbidity and mortality (P <.00001). In addition, LDL was reduced by 50%, and hs-CRP levels were significantly lower in the rosuvastatin group than in the placebo group (2.2 vs 3.5 mg/L). These results suggested that elevated hs-CRP, as least in middle aged or older patients, may identify people at higher than predicted risk for cardiovascular events than would be predicted by traditional risk factors and identify patients who should derive significant benefit from statin therapy.

Statins and Side Effects

Statin therapy was an intervention in ENHANCE, CASHMERE, and JUPITER studies. Cardiac patients often focus on the potential adverse events associated with statins, including hepatic and muscle toxicity, and the safety of combining statins with their other medications. For most patients taking statins, the incidence of liver function test (LFT) elevation >3 times the upper limit of normal (ULN) is <1% at lower statin doses and 2% to 3% at 80 mg/day. Elevated liver enzymes induced by statins are reversible after stopping the medication.¹² The rate of liver failure on statins is similar to that of the general population, and causality from statins has not been established. Liver failure occurs predominantly when statins are taken in combination with other more toxic medications.

Patients should be counseled to differentiate between myalgia and myositis. Myalgia is common and benign, whereas myositis is rare and dangerous. Cerivastatin was withdrawn from the market in 2001 in response to 31 deaths due to rhabdomyolysis in 700,000 patients. The incidence of fatal rhabdomyolyis in patients on currently marketed statins is approximately 1 per million. This can be contrasted to the risk of fatal bleeding from nonsteroidal anti-inflammatory drugs (NSAIDs), which is approximately 1 in 50,000. Some patients who experience muscle aching associated with statins improve by taking concomitant coenzyme Q, although supportive data are unclear. Muscle aching on statins is not a class effect. When a patient reports muscle aching and weakness, he or she can discontinue the drug for a week and, if symptoms are relieved, he or she can be switched to a different statin. Some patients will need to try multiple different statins but most will find an agent that they can tolerate.

Higher doses of statins are more likely to cause side effects. In some patients with multiple statin intolerances, a beneficial reduction in LDL can be achieved with longer acting statins, such as rosuvastatin and atorvastatin, given as a low dose twice a week instead of daily dosing. This often will eliminate muscle aches and can be a helpful alternative in patients who can’t tolerate daily therapy. These “tips” should be considered to increase the ability to use statin therapy in high risk patients, especially those with atherosclerosis or diabetes.

Myositis is a much more serious yet rare complication from statin therapy. Part of preventing myositis includes the avoidance of concomitant medications that increase the blood levels of statins, such as the antibiotics erythromycin and clarithromycin, antifungals, cyclosporine, fibrates such as gemfibrozil (especially with renal insufficiency), and certain antidepressants. Not all statins have interactions with these drugs; however, most of them do.

In summary, the world is facing an epidemic of obesity and diabetes. Lifestyle modification is an integral part of patient management. Calorie restriction is the best way to lose weight and should be combined with exercise to facilitate healthy weight loss and weight maintenance. In high risk patients with severe obesity, pharmacologic therapy and bariatric surgery can be very helpful. Patients’ weight loss expectations must be realistic; a 10% loss in body weight in patients with metabolic syndrome should provide an achievable target and will benefit their health significantly.

Practitioners can advocate the overall safety and tolerability as well as the efficacy of the common lipid lowering agents by counteracting misinformation when appropriate and explaining accurately the risks and benefits to their patients. Aggressive and systematic approaches to prevention are required for patients with prediabetes and diabetes, and lipid-lowering agents are critical components of a risk-lowering strategy.

References

1. Aronne LJ, Segal KR. J Clin Psychiatry. 2003;64(Suppl 8):22-29.
2. Leslie WS, et al. Q J Med. 2007;100:395-404.
3. Cheskin LJ, et al. Southern Medical Journal. 1999;92:898-904.
4. Messerli FH, et al. Am J Med. 2007;120:610-615.
5. Lewis GF. Can J Cardiol. 1995;11(suppl C):24C-28C.
6. Haffner SM, et al. N Engl J Med. 1998;339:229-234.
7. Kastelein JJP, et al. N Engl J Med. 2008;358:1431-1443.
8. O’Keefe JH, et al. Mayo Clin Proc. 2008;83:867-869.
9. Simon T, et al. Fundam Clin Pharmacol. 2004;18(1):131-138.
10. Pfizer Inc. PhRMA Web Synopsis. Protocol A2581051 29 October 2007 Final Report. Available at: http://pdf.clinicalstudyresults.org/documents/company-study_2902_0.pdf. Accessed November 29, 2008.
11. Ridker MT, et al. N Engl J Med. 2008; 359:2195-207.
12. McKenney J, et al. Am J Cardiol. 2006;97:89C-94C.

Discussion

How do the ACCORD, ADVANCE, and UKPDS consolidate your thinking about the treatment of diabetes?

Silvio E. Inzucchi, MD: I have learned that the patient’s age is not necessarily a factor; rather, it is the duration of the disease and the atherosclerotic burden. We need to appreciate those differences in practice.

Richard W. Nesto, MD: A problem arose when the ACCORD trial received publicity about whether intensive glycemic control to an approximate 6.5% A1C level reduces cardiovascular disease. The regimen used and achieving target A1C levels in 2 to 3 months is not used in clinical practice. ADVANCE provided more of a clinical expression; for example, glucose may be lowered over a 6-month or 1-year period. The key is not what intensive means; rather, the point is achieving better glycemic control.

Alan S. Brown, MD, FACC: A participant in the ACCORD trial probably would not agree to a prolonged, intensive therapy like the therapy used in that study, which was difficult for the patients. Moreover, the results have not been duplicated by another trial.

What is your advice to the average cardiologist in terms of when patients should be referred to a diabetologist?

Brown: Any patient with an A1C >10% should be referred to a diabetologist rather than the cardiologist initiating therapy. Those patients will require insulin and management. Some patients who start on oral agents and do not achieve the expected response may end up having type 1 diabetes — those patients should be referred.

Nesto: It is unfortunate that most cardiologists do not manage A1C or determine fasting glucose. Let’s measure it to make a diagnosis of diabetes in this group of patients, and if they have diabetes and it is poorly controlled, we have an obligation to either manage it ourselves or make sure the referring doctor is managing it.

Inzucchi: I have a different viewpoint for general physicians as well as cardiologists. The diabetologist offers the office infrastructure in terms of the diabetes nurse practitioner, educator, and nutritionist, and I believe he or she can do a lot more. Plus, diabetologists focus on diabetes. General physicians and cardiologists can prescribe oral agents and combinations. However, diabetologists are more facile with insulin therapies, particularly in patients with type 1 diabetes and with those on intensive treatment programs.

How do you manage the weight gain and edema that are common side effects of thiazolidinediones? If the drug is discontinued, are the adverse effects reversed?

Inzucchi: The weight gain never goes away; it is unfortunate that patients who achieve the best glycemic effect tend to gain the most weight. Patients on a combination of a glitazone and a calcium blocker, particularly imodipine, tend to experience edema. If another antihypertensive can be substituted, or lower dose of thiazolidinedione used, do so. If a diuretic is used, I prefer either spironolactone or miloride, which tends to work at a point in the distal tubule that is effective in thiazolidinedione-associated edema.

There is evidence that hyperglycemia can make platelets a bit stickier. Does the aspirin dose need to be modified for patients with diabetes?

Brown: Most endocrinologists recommend 2 baby aspirins.

Nesto: The current ADA recommendation is an 81-mg baby aspirin in any patient over the age of 40 with diabetes.

Richard H. Eckel, MD: Generally, I use a higher dose of aspirin with patients with A1C >7%, but that is purely subjective. Adequate trial data are not available.

How should fatty liver be managed?

Nesto: Fatty liver is usually associated with obesity and if you lose weight you will lose some of your hepatosteatosis. A thiazolidinedione, despite its problems, is effective at reducing intra-hepatic fat. I do not stop statins. Almost all prediabetic patients have mildly elevated liver enzymes that fluctuate because of fatty liver, and many patients will stop statins in response. If the transaminase level is not >3 times ULN, the patient should remain on statins, which may be beneficial in those patients

Eckel: I’ve used a thiazolidinedione in patients with fairly severe fatty liver and found reductions in both enzymes and fat.

Please discuss exenatide use in the ACCORD trial.

Inzucchi: There was a dramatic reduction in the primary endpoint of cardiovascular event rates and mortality rates in patients who took exenatide. Because it was a late add-on in the study, however, a survivor bias exists. It is known that GLP-1 receptors are present in the myocardium, but definitive data on cardiovascular benefits are not yet available.

Consider a patient with a low HDL, fairly advanced coronary disease, and perhaps unstable angina, who may have diabetes or impaired glucose tolerance. Is niacin an appropriate medication for these patients?

Brown: Niacin is not inappropriate in this patient. However, I believe statins should be used first, because there are considerable data on outcomes with statins in patients with prediabetes, low HDL, and high triglycerides. If the patient has a fasting glucose of 118 mg/dL, I am not sure that I would use niacin because I would be trying to avoid pushing that patient into requiring therapy for diabetes. Niacin is a reasonable alternative for a patient whose fasting glucose is 100 mg/dL.